Evidence-based evaluation | Viewpoint
Grades Without Evidence Are (Almost) Meaningless; Evidence-Based Evaluation Is Better
For many reasons--grade inflation, disparity between quality of educational institutions, confusion about what the grades actually demonstrate--the value of grades, as they are constructed now, is slipping. An emerging process using electronic portfolios produces evidence-based evaluations: richer data for better decisions during college and at graduation.
What is a grade? It’s a summative abstraction with almost no context. For decades, those in high education and those hiring college graduates made inferences from the grades that once were accepted as valid. Can we make valid inferences from grades now? As many recognize, it is hard to show a valid correlation between grades and a particular life trajectory. We may even question if the personal qualities that result in good college grades are the same necessary personal qualities for success in the world today.
No matter how we think about these issues around grades, the reality is that we now have choices for how we in higher education do final evaluations in courses. The tool is at hand to support an evaluation process that provides evidence behind the grade: the electronic portfolio (for more on electronic portfolios, see http://www.aaeebl.org). It is now possible to have a transcript with links from each grade to the work evidence behind the grade. Now, in response to the question, “but what does that grade mean?” there can be an answer.
To achieve evidence-based evaluation, an electronic portfolio (ePortfolio) must be available for students and faculty to use. And both students and faculty must know how to use ePortfolios. Right now, rarely do campuses have ePortfolios available for all students and faculty. But, even just for the value of evidence-based grading, it would be worth the investment in training time to implement ePortfolios across the campus.
At the end of each course, students would create a culminating ePortfolio of their best or most representative work during the course and a link to this culminating ePortfolio would be included in the official grade for that course.
I have often agonized over final grades myself, disturbed that the one letter abstraction leaves out so much: An “A” for a student who started the semester poorly but caught on, got inspired and engaged, and gradually pulled herself up to solid achievement is not really the same as an “A” for a student who started out skilled, changed little, but still earned an “A.” I want to say so much more about each student than a simple choice among 5 letters.
An Even More Exciting Possibility
If an institution achieves the move to evidence-based, contextualized evaluation, meaning that students and faculty are actively and purposefully using ePortfolios, then another possibility presents itself: outside eportfolio assessment. The instructor in the course can share the assessment of the student’s culminating ePortfolio with another instructor or even with another student. I heard about this process at Clemson University and have wondered if, as another variation on this assessment process, the teacher in the course has to be involved in final evaluation at all.
In the current system of grading, only the teacher has the knowledge of her students sufficient to grade. But in an evidence-based evaluation system, the evidence should speak for itself and an outside group of evaluators, reviewing the culminating ePortfolio could agree on a final grade.
The reason for using outside evaluators is not just to save the teacher work or to spare her the final-grade agony I experienced and which I am sure other faculty members experience, but for objectivity and for another very subtle but crucial value: When the final grade is determined by an outside team of evaluators, the teacher then becomes an ally preparing students for their final evaluation.
The teacher will still assess (assess and evaluate are not the same) work during the semester and that assessment will figure in the calculations for the transcript, but the outside team does the final evaluation. Students could then, to the extent this applies in each case, feel freed of the constant need to “please the teacher” to get a good grade in the course. They could focus instead on creating the best possible portfolio for unknown evaluators.
A Final Twist
As a next step in “going ePortfolio,” once students are creating a culminating ePortfolio on which to be evaluated by an outside group as the operational grading process on your “ePortfolio-ized” campus, your campus could consider changing the role of faculty to preparing students at the beginning of each semester with sufficient information for them to solve problems in groups, using ePortfolios to collect evidence, and then letting them work semi-independently for a number of weeks.
The “semi-independently” would be different in scope in the first years of the undergraduate experience than later in major programs or in graduate school, but the process could remain essentially the same.
Undergraduate courses would still start with “informational learning” (Robert Kegan) but would then, with the ePortfolio phase of the course, move to “transformational learning,” during which time each cohort of students entering this course would change the form of their learning in that field.
If faculty members are freed from lecturing each day of the semester and from the final grading process, while still assuring that students meet the learning goals of the course, their (the faculty members’) psychological workload may be experienced as lighter. For adjunct faculty, the pressure to give good grades so their student evaluations would be high and they would be more likely to be hired for another semester, would also be lessened under this model. Each institution, as well, could feel more confident in the objectivity of student evaluations.
The answer to “accountability” is not more standardized testing but evidence-based evaluation. Standardized testing probably doesn’t improve learning, but evidence-based evaluation probably does.